The IRF5 gene has been associated with risk for lupus. These findings demonstrate that the lupus-associated polymorphisms in the IRF5 gene have wide-reaching effects on B-cell responses to infection. The gene sets that were enriched in the risk haplotypes included interferon-related sets, which is encouraging considering that the genotype being examined is IRF5. Multiple gene sets that are related to lupus were enriched in the cells with the risk haplotypes, including IFNα sets, interleukin- (IL-) 6, inflammation, proliferation, and monocyte and dendritic cell genes, in addition to the SLE-related interferon gene set. The finding that these gene sets are the most strongly enriched in the risk haplotype indicates that the IRF5 risk haplotype has a strong influence on interferon signaling and inflammation, processes that are at the core of SLE. The finding that the most enriched gene sets were associated with interferon and lupus also indicates that these results are unlikely to be false positives obtained by chance, since the variable being studied is an interferon-affecting gene.
The gene set enrichment analysis techniques that were used are valuable because they identify not only individual genes, but also how strongly pathways that include those genes and the interactions between them are affected by the experimental conditions. This allows a much broader look into gene networks than looking only at individual genes. These studies point to the IRF5 risk haplotype having a wide influence on interferon and inflammation. The results identify targets for future investigation into the function of the IRF5 polymorphisms as well as other genetic influences on lupus.
As was the case with the gene set enrichment analysis, the identification of the interferon and Toll-like receptor pathways through Illumina pathway analysis suggests that the results are robust, as these are pathways that would be expected to be modulated by the underlying IRF5
haplotype of the donor. Interferon alpha is an extremely important cytokine in lupus [53
]. These studies suggest that the interferon alpha pathway is strongly affected by genetic variation in the IRF5
gene, and show multiple genes that could potentially be targets for understanding interferon in lupus or potential therapeutic targets. Toll-like receptors are involved in response to infection through the recognition of pathogen-associated molecular patterns. Additionally, Toll-like receptors are important in the pathogenesis of lupus. They are capable of recognizing endogenous nucleic acids in the context of immune complexes found in lupus patient sera, thereby stimulating dendritic cell maturation and interferon alpha production [25
], a process which also involves IRF5
is a very interesting transcriptional regulator in that it acts as both an activator when homodimerized and blocks activation when heterodimerized with IRF7
]. This mechanism of action may help to explain how some of these pathways can exhibit relative upregulation or downregulation depending on the other conditions in the cell.
One of the more unexpected findings of this study was the modulation of the B-cell receptor pathway by the IRF5
haplotype. The B-cell receptor is important in the recognition of antigen and the survival, maturation, and proliferation of B cells. B cells produce the autoantibodies involved in lupus, as well as being important for antigen processing and presentation and T-cell activation. Differences in the activation threshold or other effects that may be seen with altered B-cell receptor gene expression may be very important to breaking self-tolerance or other aspects of B cell biology involved in SLE. Of particular interest in this respect is the recent finding that IRF5
controls antibody class switching to IgG2A, allowing lupus-like autoimmunity in mice [63
]. The B-cell receptor and Toll-like receptor pathways are involved in antibody class switching, and the genes that were modulated by IRF5
variation in this study could represent mechanisms through which the IRF5
risk haplotype may contribute to class switching or other similar variations in humans.
The interferon response signature has been identified as a common feature in lupus. These studies examined how polymorphisms in the IRF5 gene affected the interferon response signature in both patients and controls. Interestingly, SLE patients with the risk haplotype demonstrated an interferon signature in both the infected and uninfected cells, while a strong interferon response was not found in the patients without the risk haplotype even when stimulated by EBV exposure. The controls with the risk haplotype lacked the interferon response signature in the basal state, but developed it after exposure to EBV, as would be expected. These findings suggest that the IRF5 risk haplotype is integral for the interferon response signature in both patients and controls. They also indicate that other factors contribute to a basal interferon response in lupus patients, since the IRF5 risk haplotype was not sufficient for the response signature to be present in the unstimulated control cells, as it was in the patients with the risk haplotype.
Cells were infected with EBV for two reasons. The first was to identify differences in gene expression patterns when cells were stimulated with a biologically relevant trigger for interferon production. The second reason was to identify areas that may start to explain the differences in EBV infection and response in lupus. Gene expression was examined for genes in the three pathways found to be significant by ingenuity analysis. In several cases, (IFITM1, IFNAR2, LY96, PIK3CA, NFAT5, and GSK3B) the baseline level of gene expression was higher in the risk cells, but after EBV infection, the gene expression was comparatively increased in the protective cells. In other genes, including CD79A, CD79B, STAT1, MyD88, and Fos, expression was lower in the risk cells but the difference diminished or reversed after EBV infection. Expression of one gene, TNFAIP3, was lower in the risk than in the protective haplotype subject unstimulated cells and was comparatively diminished further after EBV infection. These differences suggest several areas of investigation to understand differences in B cell biology in lupus and show that the IRF5 haplotype affects multiple genes related to EBV infection and response.
Although a detailed analysis of each gene involved in these pathways is beyond the scope of this paper, the genes with expression differences between the risk and protective haplotypes are suggestive in several instances. One of the genes identified with promise to affect lupus is TNFAIP3
. This gene is a transcription factor that is produced in response to inflammation. It has been shown to be critical to limiting inflammation by terminating NFκ
B responses [64
].Variants have recently been associated with risk for lupus and other autoimmune and inflammatory diseases [65
], and it is often suppressed in tumors, especially lymphomas [66
]. Other promising genes identified by these experiments include STAT4
, and IFNAR2
, which are all involved in the response to interferon, and several B cell signaling genes, including NFAT5
, and NFκB2
Although EBV was used in part to simulate an infected state in B cells, EBV itself could be involved in the etiology of lupus by affecting several pathways. The three pathways identified here are all involved in EBV infection. EBV may stimulate these pathways through several mechanisms, including both infection and binding of virions to the receptors involved in these pathways. Although the effect of EBV infection on differential gene expression was somewhat variable, for many of the genes examined in this study there was overexpression in the risk cells, which subsequently diminished after EBV infection. This pattern, as well as that seen with the interferon response signature, suggests that the IRF5 risk haplotype makes these cells appear more activated in the resting state. Because of this heightened activation state, there is less difference in the response to EBV infection in the risk cells, with the nonrisk cells often catching up to or passing the risk cells in expression of several genes following viral infection. An activated basal state would be likely to promote inappropriate cellular responses and possibly heightened sensitivity to self-antigens, including those recognized by TLRs.
These findings identify several key pathways that are affected by the IRF5 risk haplotype and are involved in the B cell response to antigen stimulation and viral infection. Many of the genes involved in these pathways have definite potential to alter the response to EBV infection and affect the development of lupus. These merit further investigation. Since all of these pathways are likely to be involved in the development of lupus, further comparison of these pathways in other cell types such as plasmacytoid dendritic cells will be beneficial to understanding the origins and pathogenesis of lupus. It will also be beneficial to examine more closely the role of EBV in regulating expression of these genes, through the use of EBV mutants, and to dissect the role of IRF5 in each pathway and gene set identified.